We have continued to develop, implement, and apply simulation methods to expedite conformational searching in the prediction of peptide structure (Proteins 2004), to understand the interactions of proteins with solvent (Journal of Physical Chemistry B, in press), to model proteins by homology, and to refine such models. With NCI researchers, we proposed that Heat Shock Protein 90, an important target for the killing of cancer cells, catalyzes a particular conformational change of geldanamycin (Nature 2003; Chemistry and Biology 2004). We continue to work on the design of potent HSP90 inhibitors. Another paper is in preparation describing additional modeling that rationalizes experimental data and proposes an explanation of HSP90 inhibition. The challenges faced in modeling protein-ligand interactions were discussed in a book chapter (Humana Press, in press). In collaboration with NIMH colleagues, quantum chemical calculations were done in search of more potent and selective positron emission tomography (PET) ligands for receptors in the brain (manuscript in preparation). Quantum mechanical calculations have been used to study proton transfer in bacteriorhodopsin (J. Am. Chem. Soc., 2004) and the mechanism of adenylyl cyclase (J. Phys. Chem. B., 2004). A model of protein GP1b-beta was built for a study with NICHD scientists that sheds light on mutations of this protein that are associated with the bleeding disorder Bernard-Soulier syndrome(Thrombosis and Haemostasis 2004). A model of integrin alphaIIb-beta3, also associated with a bleeding disorder, was published (Journal of Biological Chemistry 2004) and deposited in the Protein Data Bank. Two additional manuscripts are in preparation that use homology models to describe the effects of specific mutations on the functioning of proteins. In collaboration with NIDDK, we are characterizing the chemistry, thermodynamics, and spectroscopic properties of carbon nanotubes and their interactions with confined organic molecules. As an idealized model of biological systems such as transmebrane channels, these systems may find utility in nanomedicine. We have also continued to study peptide dynamics using Langevin simulations and an implicit solvent model. A manuscript summarizing this work is in preparation. Work continues on a collaborative study of N-acetyltransferase. QM/MM dynamics simulations of the proton transfer in a carbon nanotube are in progress. Molecular modeling studies on the A-beta amyloid probes are being carried out to establish a structure-activity relationship. Quantum chemical calculations are also being carried out to quantify the interaction energy between propanediol and water molecules.